Skip to main content

Table 3 Posterior median (and 95% CrI) and bias (and width of 95% CrI) for common Ï„2

From: Pattern-mixture model in network meta-analysis of binary missing outcome data: one-stage or two-stage approach?

small Ï„2

moderate MOD

large MOD

trial size

frequency

one-stage

two-stage

one-stage

two-stage

small

low

0.13 (6 × 10− 3, 2.59)

0.11 (2.58)

0.10 (6 × 10− 3, 1.14)

0.08 (1.13)

0.13 (6 × 10− 3, 3.06)

0.11 (3.06)

0.10 (6 × 10− 3, 1.20)

0.08 (1.19)

moderate

low

0.07 (4 × 10− 3, 0.70)

0.05 (0.70)

0.07 (5 × 10− 3, 0.64)

0.05 (0.63)

0.08 (5 × 10− 3, 0.97)

0.07 (0.96)

0.08 (5 × 10− 3, 0.82)

0.06 (0.81)

small

frequent

0.09 (5 × 10− 3, 0.94)

0.07 (0.93)

0.08 (5 × 10− 3, 0.83)

0.06 (0.82)

0.09 (5 × 10− 3, 1.04)

0.07 (1.03)

0.09 (6 × 10− 3, 0.91)

0.06 (0.90)

moderate

frequent

0.04 (3 × 10− 3, 0.31)

0.02 (0.31)

0.04 (4 × 10− 3, 0.32)

0.02 (0.31)

0.05 (4 × 10− 3, 0.40)

0.03 (0.40)

0.05 (4 × 10− 3, 0.42)

0.03 (0.42)

substantial Ï„2

moderate MOD

large MOD

trial size

frequency

one-stage

two-stage

one-stage

two-stage

small

low

0.13 (6 × 10− 3, 2.52)

0.05 (2.51)

0.10 (6 × 10− 3, 1.16)

0.02 (1.15)

0.14 (6 × 10− 3, 3.29)

0.06 (3.29)

0.10 (6 × 10− 3, 1.21)

0.02 (1.21)

moderate

low

0.09 (5 × 10− 3, 1.05)

0.01 (1.05)

0.08 (5 × 10− 3, 0.86)

0.004 (0.85)

0.09 (5 × 10− 3, 1.17)

0.02 (1.17)

0.08 (5 × 10− 3, 0.90)

0.01 (0.89)

small

frequent

0.10 (5 × 10− 3, 1.15)

0.02 (1.14)

0.09 (6 × 10− 3, 0.95)

0.01 (0.95)

0.10 (5 × 10− 3, 1.26)

0.02 (1.25)

0.09 (6 × 10− 3, 1.01)

0.01 (1.01)

moderate

frequent

0.06 (4 × 10− 3, 0.57)

−0.02 (0.57)

0.05 (4 × 10− 3, 0.56)

− 0.02 (0.56)

0.06 (4 × 10− 3, 0.58)

− 0.02 (0.57)

0.06 (4 × 10− 3, 0.56)

− 0.02 (0.56)

  1. MOD Missing outcome data
  2. Posterior median and 95% CrI (in parenthesis) are provided in the first line of every cell, followed by bias and width of 95% CrI (in parenthesis) in the second line of every cell